I’ve been using Max since its inception, but only recently needed to dabble in Reactor.  I’m trying to animate kinetic art pieces.  Things like wind chimes, mobiles and other sculptures where all the movement is based on wind and resistance.  I’m really hoping Reactor can do the job because doing it manually can get quite tedious.  Below is just one issue I’ve come up with.  I’m hoping if this can be solved, other projects I’d like to persue can be accomplished.

The set up is pretty simple.  Its just a ball rolling around in a bowl.  That works great.  Problems begin when I want to animate the bowl.  I’ve tweaked all the properties of the objects but there’s always some virtual crack as soon as the bowl begins to move.  Lets say the bowl is floating in water, so its slowly rotating and shifting.  When this occurs, the ball falls through the bowl.  Its only when the bowl moves though!  I find this happening in a lot of simulations when you animate the object acting as the ground .  All the collisions get thrown off.  Is this a limitation of Reactor or am I missing something very simple?  I really hope its just a bone headed miss on my part. 

Please help.

I don’t think its a bonehead move on your part.  Most physics tools can get finicky when it comes to something as complicated as what you are talking about.  I’ll have to play with it a little to see if I can make it work.  Likely this will come down to sampling rates and possibly scale issues but hard to say at the moment.  One great thing about reactor in relation to wind driven effects is that reactor actually does wind occlusion which I have found rare in dynamics systems.

Shawn Hendriks
Technical Marketing and Video Production Sr. Manager
Autodesk Media and Entertainment
My AREA Blog
My twitter feed

Well, I just quickly rebuilt the scene and it worked this time.  The ball rolls around in the bowl as its swaying just like its supposed to now.  Go figure.  Thanks for the input Shawn.

How’s this for a wrinkle on the ball in a bowl?  I carved out a track in the bowl to see if the ball will roll around inside this trench.  It rolls just fine on the way down, but as soon as the bowl rotates up again, the ball keeps going right on through.  See the movie attachment.  It worked fine inside the bowl but not in a track running around the rim.  Wacky????

	Attachment
6921_3JabOm2xLUBT5L46ex2z.mov

If you’re only working with rigid body simulations (no rope, cloth, etc) try using the Havok3 version instead of the Havok1 (changeable under the first rollout in the reactor utilities menu)

I’ve found the Havok3 solver to be quite a bit more stable for things rolling/rotating (which tend very much to fall through the geometry they’re on...)

SJ

Yup that did it.  Havok 3 worked great,